Acute exercise-induced catecholaminergic responses after 16 weeks of community-based exercise training in early-stage breast cancer survivors

Abstract

BACKGROUND

Breast cancer survivors (BCS) demonstrate attenuated immune cell mobilization following acute exercise, with partial restoration following exercise training. Epinephrine (EPI) and norepinephrine (NE) are responsive to exercise-stress and directly regulate immune cell function, indicating a potential role in this restorative process. Similar attenuations in catecholaminergic signaling have been reported in BCS post-exercise; however, it is unknown whether this is maintained within a trained state. We hypothesized that compared to non-cancer controls (CON), acute exercise would induce an attenuated catecholaminergic response in untrained BCS, which would be recovered to levels similar to CON after training.

AIM

To compare acute exercise-induced catecholaminergic responses between BCS and CON before (PRE) and after (POST) completing a community-based exercise intervention.

METHODS

Thirteen BCS (age: 56 ± 2 years, body fat: 39.7% ± 1.3%) and 13 CON (age: 56 ± 2 years, body fat: 41.2% ± 1.7%) performed 45 minutes of intermittent cycling at 60% peak power output PRE and POST 16 weeks of community-based exercise training. Blood samples were collected at baseline (BASE), immediately (0 hour), and 1-hour (1 hour) post-exercise for assessment of the acute EPI and NE response. Separate linear mixed models were used for PRE and POST EPI and NE assessment.

RESULTS

At PRE, both BCS and CON demonstrated increases in EPI (+87.4 pg∙mL^-1, P < 0.001) and NE (+1295 pg∙mL^-1, P < 0.001) at 0 hour, with no group differences. At POST, group differences in NE initiation (0 hour-BASE) were not statistically significant (-544.9 pg∙mL^-1, P = 0.115, g = 0.92), despite divergent responses between BCS (+28%, P = 0.175, g = 0.36) and CON (-13%, P = 0.377, g = 0.23). No group differences were observed for NE recovery (1 hour-0 hour) nor for EPI initiation or recovery.

CONCLUSION

BCS and CON present with similar exercise-induced catecholaminergic responses regardless of training, suggesting an alternative mechanism may have made a greater contribution to the training-induced immune cell revival previously observed.

Key Words: Autonomic function; Breast neoplasms; Cancer survivorship; Cardiovascular stress response; Catecholamines; Exercise therapy; Sympathetic nervous system

Core Tip: Exercise training partially restored immune cell responses previously diminished within untrained breast cancer survivors (BCS) immediately post-exercise. Epinephrine and norepinephrine are exercise-responsive and regulate immune cell function, suggesting a role in this training-induced restoration. Acute exercise induces similar BCS catecholaminergic attenuation; however, it is unclear whether this is sustained post-training. In this study, we aimed to compare acute exercise-induced catecholaminergic responses between BCS and non-cancer controls (CON) before and after a training intervention. BCS and CON presented with similar exercise-induced catecholaminergic responses regardless of training, suggesting the immune cell revival previously observed may be regulated by alternative mechanisms.

INTRODUCTION

Breast cancer (BC) is one of the most common cancer types in the United States, accounting for approximately 30% of all new female cancers per year; yet, developments in early detection and preventative treatments have maintained a 5-year overall survival rate of 91%[1]. Despite these advancements, breast cancer survivors (BCS) face a reduced quality of life due to the side effects of treatment. Amongst these are the combined impact of aging and treatment-related immunosuppression[2-5]. The chronic activation of adrenergic receptors (ARs), which are expressed on the surface of immune cells, may contribute to this immunosuppressive state[6,7]. Epinephrine (EPI) and norepinephrine (NE), catecholamines produced by the stress-activated sympathoadrenal system, are well-known for regulating the fight-or-flight response via AR binding mechanisms[6]. Yet, despite their well-characterized role in stress physiology, the catecholaminergic role in cancer biology remains a mechanistic conundrum[8]. For example, chronically elevated catecholamine levels have been linked to enhanced BC tumorigenesis[9,10]; positioning the adrenergic system as a viable target for anti-cancer treatments[11,12] which has resulted in high rates of cardiovascular dysautonomia among BCS[13,14]. Alternatively, exercise-induced catecholamines appear to exert immunostimulatory effects, enhancing immune cell mobilization and activation[15,16], which may support anti-tumor surveillance and control[17].

Weakened cellular immunity and chronic inflammation are significant concerns for BCS, as they contribute to an immunosuppressed state and elevated risk of cancer recurrence[5]. Exercise training, a non-pharmacological lifestyle factor, has been linked to enhancements in physical function, cardiorespiratory fitness, body composition, and inflammatory status in both BC patients and survivors[18,19]. Within BCS, the immunomodulatory benefits of exercise seem to be altered, resulting in differential observations compared to cancer-free individuals. Our lab has previously observed an approximately 50% attenuation of exercise-induced EPI (vs matched non-cancer controls (CON)[20]), with similar findings reported in prostate cancer (PCa)[21]. Such an acute exercise-induced adrenergic alteration may contribute to metabolic inefficiencies - characterized by lower glucose metabolism, increased fat oxidation, and reductions in acute exercise capacity[20,22] - as well as attenuated immune responses, marked by reduced mobilization and activation across multiple immune cell types[23,24].

We hypothesize that such exercise-induced responses could arise from adrenergic dysfunction as a result of autonomic abnormalities, which have been previously observed in BCS[13,14,20]. Characterized by the downregulation of AR density and responsiveness[7], adrenergic dysfunction is often a consequence of chronically elevated sympathoadrenal activity, which may be accompanied by greater basal levels of catecholamines in BCS[20]. One consequence of this prolonged agonistic exposure is the suppression of lymphocytic activity[6,7]. Previously, our group demonstrated an acute attenuation of immune cell mobilization accompanied by functional alterations in BCS mucosal-associated invariant T-cells (MAIT)[24] and neutrophils[23]. Interestingly, following a 16-week community-based exercise program, there was partial rescue of these phenotypes in BCS. This recovery could be mediated, at least in part, by training-induced sympathoadrenal adaptations[23,24]. However, the notion that exercise can ‘train’ the sympathoadrenal network to enhance functional immune cell activation requires further investigation[7].

Despite evidence of a robust (+120%) catecholaminergic response following acute exercise[17,25], EPI and NE remained unaltered within trained cancer[26] and non-cancer populations[27]. Although unclear, this discrepancy may elucidate a methodological limitation in using resting samples to explore training-induced differences in catecholaminergic signaling. This approach fails to provide an acute stimulus (e.g., exercise) to induce a sympathoadrenal response necessary for the assessment of training-induced response adaptations[26,27]. To our knowledge, the direct examination of basal and exercise-induced catecholaminergic responses within a naïve and trained state has evaded investigation within BCS. Such an approach is needed to determine how, or even if, the catecholaminergic response may underlie the immune improvements previously observed in trained BCS[23,24]. Therefore, this study has two aims: (1) To determine if an altered (a) basal and/or (b) exercise-induced adrenergic response explains the previously reported attenuation of exercise-induced MAIT cell[24] and neutrophil[23] mobilization and activity prior to a 16-week community based exercise intervention; and (2) To explore if this exercise training plan results in catecholaminergic normalization relative to CON. We hypothesize that within a naïve state, BCS would present with a blunted adrenergic response immediately post-exercise compared to CON. However, when reassessed within a trained state, we anticipate that exercise-induced adrenergic responses in BCS would recover to levels similar to CON. We sought to explore this by comparing plasma EPI and NE responses in BCS and CON who underwent an acute exercise stimulus before (PRE) and after (POST) a 16-week exercise intervention designed to meet current exercise oncology guidelines[28].

MATERIALS AND METHODS

This investigation involved a pre-planned, secondary analysis of exercise-induced catecholaminergic responses in BCS compared to CON. The study design and methodology have been previously described and consisted of mostly the same women in our previous studies[23,24], who were included from a more extensive investigation registered with the Get REAL and HEEL Research Program (ClinicalTrials.gov ID: NCT03760536). Briefly, thirteen BCS (previously diagnosed with stage I-III BC with no more than one year since final treatment) and CON were matched for age and body fat percentage. Participants were inactive, defined as failing to meet American College of Sports Medicine guidelines[29]. All study procedures were approved by the Protocol Review Committee at the Lineberger Comprehensive Cancer Center (LCCC1630: Get REAL and HEEL Research Program) and by the Institutional Review Board of UNC-Chapel Hill (IRB #16-3284). Prior to study enrollment, all participants provided written informed consent.

Study design

This study utilized a unique design, with acute and chronic exercise included for the assessment of circulating catecholamines, along with the use of CON as a referent group for the optimal response to training. Here, all participants performed an acute, intermittent cycling exercise bout PRE and POST a 16-week community-based aerobic and resistance training program, consisting of approximately 1 hour of strength and aerobic exercise sessions performed three days per week[30].

Preliminary assessments, familiarization, and cardiopulmonary exercise testing

Visits 1-3 for all participants were performed as previously described[23,24,30]. Briefly, during visit 1 participants were medically cleared and familiarized with the cardiopulmonary exercise test (CPET) protocol. The CPET involved a continuous ramp (15 W∙minute^-1) cycling protocol, which was terminated once 75% of the estimated heart rate reserve was met. Visit 2 included a body composition assessment via dual-energy X-ray absorptiometry (DXA: Discovery W, Hologic Inc., Bedford, MA, United States) and the completion of a maximal effort CPET to volitional exhaustion. Expired respiratory gases were analyzed using a metabolic system (TrueMax 2400, Parvo Medics, Salt Lake City, UT, United States) and peak power output (PPO) was recorded as the maximal wattage completed in the last minute prior to test termination. Visit 3 was performed within 7 days of maximal CPET completion. Participants reported to the laboratory between 0700 AM and 1000 AM in a hydrated state and having refrained from caffeine (8 hours) as well as alcohol and strenuous exercise (24 hours). After 10 minutes of supine rest, an indwelling venous catheter was inserted and a 6 mL ethylenediamine tetraacetic acid (K2) blood sample was collected (BASE). After a brief warm-up, participants completed 10 intervals of cycling consisting of 3 minutes at 60% PPO followed by 1.5 minutes of passive recovery. This exercise intensity was used due to its (1) Eclipse of the intensity threshold required to induce a sympathoadrenal response[27,31]; and (2) Previously reported feasibility amongst cancer populations[21]. Both heart rate (HR) and rating of perceived exertion (RPE) were determined in the last 30 seconds of each stage. Additional blood samples were obtained immediately (≤ 30 seconds) following exercise cessation (0 hour) and after 1 hour of seated recovery (1 hour) with ad libitum water access. All blood samples were placed on ice until the cessation of data collection procedures and then immediately centrifuged at 200 × g for 10 minutes at room temperature. This step was completed due to the concomitant collection of peripheral blood mononuclear cells, which were considered a primary outcome of other studies which included this BCS cohort[23,24]. The resulting supernatant was immediately aliquoted and stored at -80 °C for future analysis.

Exercise training intervention

Participants completed a 16-week community-based exercise intervention designed to meet current exercise oncology guidelines (ClinicalTrials.gov ID: NCT03760536)[28,30]. Exercise sessions were performed within small groups over a duration of approximately 1 hour, 3 days per week, and consisted of aerobic and resistance exercises administered by trained exercise specialists. To create an exercise training dose-response, duration (aerobic only) and intensity (Borg RPE scale) were monitored and progressively increased across the 16-week training period, whereas resistance training duration was maintained at approximately 30 minutes per session. Participants were always supervised and continuously monitored by training staff for adherence and compliance with the prescribed training intervention. Exercise adherence was defined as the percentage of days attended compared to prescribed (48 total), whereas exercise compliance consisted of completing ≥ 80% of the prescribed duration (minutes; aerobic exercise) or volume (sets by repetitions; resistance exercise) at the prescribed intensity (RPE) relative to the total sessions prescribed[23,24,30]. Following the completion of the 16-week exercise intervention, the same BASE assessments (visits 1, 2, and 3) were repeated.

Biospecimen analysis

Prior to centrifugation, complete blood cell counts were performed in duplicate using an automated hematology analyzer (XP-300™, Sysmex Corporation, Kobe, Japan) for the measurement of lymphocyte, MAIT, and neutrophil cell counts as previously reported[23,24]. For the quantitative determination of EPI and NE, a competitive-binding ELISA immunoassay (17-BCTHU-E02.1, ALPCO, Salem, NH, United States) was used. Manufacturer-reported limits of detection were 10 pg∙mL^-1 for EPI and 36 pg∙mL^-1 for NE, with average recovery rates of 105% and 87%, respectively. Prior to analysis, samples were thawed at room temperature and spun for approximately 5 minutes at 16000 × g to remove any remaining red blood cells. The resulting supernatant was measured in duplicate according to manufacturer instructions and analyzed via absorbance-based detection (ChroMate 4300 Microplate Reader, Awareness Technology, Inc., Palm City, FL, United States). The inter-assay coefficients of variation were 5.16% and 5.33% and the intra-assay coefficients of variation were 1.41% and 0.61% for EPI and NE, respectively. The standard curve for EPI (0-200 ng∙mL^-1) and NE (0-1000 ng∙mL^-1) had correlation coefficients of r² = 0.99 and r² = 0.98, respectively.

Statistical analysis

Statistical analysis was performed using Jamovi (version 2.4; the Jamovi project, Sydney, Australia) and data was visualized in GraphPad Prism (Version 10, GraphPad Software, Boston, MA, United States). A sample size power calculation for the parent trial was performed a priori for the primary outcome of VO₂peak, using an effect size (Cohen’s d) of 0.83, α of 0.05, and β of 80%[30]. The current analysis evaluated a different, secondary outcome for which the study may not be specifically powered. Statistical significance for main effects was set at P = 0.05, and P = 0.10 for borderline significant interactions. Effect sizes for group differences were estimated using Hedge’s g, such that 0.16, 0.38, and 0.76 represent small, medium, and large differences, respectively[32]. Our statistical analysis follows an approach utilized previously[24]. Results are presented as mean ± SD, with model estimates for mean differences expressed relative to BASE (for time effects), PRE (for phase effects), or CON (for group effects), along with 95% confidence intervals. Baseline participant characteristics at PRE were determined using independent samples t-tests and χ² tests. Training-induced changes were assessed using a group (BCS vs CON) × training (PRE vs POST) linear mixed model approach. To determine if adrenergic alterations underlie differences in exercise performance (Aim 1), all PRE acute exercise-induced EPI and NE responses were evaluated using separate linear mixed models, with group (BCS vs CON) and time (BASE, 0 hour, 1 hour) as fixed factors, and subjects as a random effect. Time-by-group interactions were resolved using simple effects to compare groups at each time point. To explore training-induced catecholaminergic adaptations (Aim 2), EPI and NE responses were separated into two phases: Initiation (0 hour-BASE) and recovery (1 hour-0 hour). These measures were taken to mitigate the risk of false positive interpretations which are enhanced within 3-way interaction analyses[33]. Within each separate linear mixed model, fixed factors were group (BCS vs CON) and training phase (PRE vs POST), with subjects included as a random effect. Simple effects were again used to resolve any training-by-group interactions (e.g., group comparisons for pre-post intervention response). For both Aim 1 and 2 statistical approaches, post-hoc pairwise comparisons were performed using Tukey’s honest significant differences (HSD) test. Additionally, total phase-specific responses were quantified using the trapezoidal calculation for area under the curve (AUC), which were pursued as an unplanned post hoc analysis to confirm the effects reported in our Aim 2 analysis. Lastly, Pearson correlation matrices were conducted as an exploratory measure to assess the association between the catecholaminergic responses observed and immune cell counts previously reported within this cohort[23,24]. Correlations were described using coefficient stratifications of negligible (0.00-0.10), weak (0.10-0.39), moderate (0.40-0.69), strong (0.70-0.89), and very strong (0.90-1.00) as well as with coefficients of determination (r²) as previously described[34].

RESULTS

Participants

No differences in BASE PRE measurements were reported between pre- and post-menopausal women in either group, who were matched for age and body composition. Both groups were comprised of primarily college-educated white, non-Hispanic females who were currently married and employed (Table 1). At POST, participant characteristics remained unchanged except for lean mass% (P = 0.030) which increased in BCS (+1.8% ± 0.8%; P = 0.019) compared to a non-significant decrease in CON (-1.1% ± 0.8%; P = 0.825). BCS had mostly early-stage diagnoses, and all underwent surgery prior to secondary treatment initiation [radiation (46%), chemotherapy (23%), or both (31%)]. At enrollment, BCS were approximately 3 months post-primary treatment completion, with the majority identified as post-menopausal (Table 2). Evidence of cancer or treatment-induced adrenergic dysfunction was not observed, with similar measurements in basal EPI (P = 0.096) and NE (P = 0.543) reported between BCS and CON at PRE (Table 1). Of the 48 prescribed training days, no group differences in adherence were identified between BCS (72% ± 19%) and CON (69% ± 19%). However, exercise compliance was greater in CON for both aerobic (CON: 67% ± 19%, BCS: 54% ± 21%, P = 0.027) and strength training (CON: 38% ± 6%, BCS: 29% ± 10%, P = 0.019).

Table 1 Participant characteristics before and after a 16-week community-based exercise training program, mean (SD) or n (%).

	BCS (n = 13)		CON (n = 13)		P values
	PRE	POST	PRE	POST	G × P	Group	Phase
Age (year)	57 (9)		58 (7)			0.658
Race (white)	13 (100)		13 (100)			1.000
Ethnicity (non-Hispanic)	13 (100)		12 (92.3)			0.308
Education (college degree)	12 (92.3)		13 (100)			0.288
Employment status (employed)	8 (61.5)		12 (92.3)			0.135
Marital status (married)	8 (61.5)		8 (61.5)			0.790
Height (cm)	164.5 (9.0)		162.7 (5.5)			0.724
Body mass (kg)	69.8 (8.6)	70.0 (8.3)	72.3 (19.5)	72.7 (20.8)	0.737	0.751	0.549
Lean mass (kg)	38.7 (3.7)	40.0 (3.0)	41.2 (9.4)	40.7 (10.6)	0.083	0.680	0.388
Body fat (%)	40.9 (7.2)	39.2 (6.4)	39.2 (4.4)	40.4 (5.2)	0.053	0.949	0.546
Lean mass (%)	55.9 (6.8)	57.7 (6.0)	57.6 (3.9)	56.5 (4.7)	0.030a	0.947	0.625
Epinephrine (pg∙mL-1)	12.5 (11.6)	15.0 (15.9)	23.5 (12.0)	20.3 (30.2)	0.999	0.447	0.899
Norepinephrine (pg∙mL-1)	565.6 (295.2)	467.4 (305.9)	653.6 (403.4)	679.0 (322.2)	0.162	0.671	0.610

^aP < 0.05, interaction effect [breast cancer survivor (BCS) vs non-cancer control (CON)].

Independent samples t-tests and χ² tests compared characteristic differences between BCS and CON. BCS: Breast cancer survivor; CON: Non-cancer control; PRE: Pre-training; POST: Post-training; G × P: Group × phase interaction.

Table 2 Clinical characteristics of breast cancer survivors, mean (SD) or n (%).

	BCS (n = 13)
Time from last treatment (day)	82 (90)
Time from surgery (day)	224 (115)
Time from radiation (day)	62 (78)
Time from chemotherapy (day)	152 (100)
Stage
I	5 (38)
II	6 (46)
III	2 (15)
Tumor status
ER+ positive	12 (92)
PR+ positive	12 (92)
HER2+ positive	4 (31)
ER-/PR-/HER2-	1 (8)
Pre-menopausal	5 (38)
Post-menopausal	8 (62)

BCS: Breast cancer survivor; ER: Estrogen-receptor; PR: Progesterone-receptor; HER2: Human epidermal growth factor receptor 2.

Exercise testing

Both CPET (visit 2) and acute intermittent exercise trial (visit 3) physiological responses are presented in Table 3. Average HR and RPE during PRE and POST acute exercise did not differ between groups. During acute exercise, BCS reached 85% peak HR for both PRE and POST while CON reached 86% and 92%, respectively.

Table 3 Physiological responses to VO₂peak test and acute, intermittent exercise before and after training, mean (SD).

	BCS (n = 13)		CON (n = 13)		P values
	PRE	POST	PRE	POST	G × P	Group	Phase	g
CPET
VO2peak (mL∙kg-1∙minute-1)	21.4 (6.6)	22.3 (5.6)	22.5 (2.8)	23.6 (4.5)	0.860	0.523	0.164	-0.04
Peak HR (bpm)	160 (18)	160 (17)	157 (10)	158 (12)	0.830	0.709	0.910	-0.07
Peak power output (W)	119 (34)	137 (40)	129 (21)	136 (24)	0.050a	0.713	< 0.001b	0.34
Duration (second)	596 (127)	669 (142)	617 (85)	647 (154)	0.236	0.928	0.006b	0.31
Acute Exercise Trial
Resistance (W)	74 (20)	83 (24)	77 (16)	82 (82)	0.194	0.848	< 0.001b	0.07
Average RPE	14 (3)	13 (1)	13 (1)	14 (1)	0.070	0.709	0.588	-0.97
average HR (bpm)	136 (8)	136 (7)	135 (8)	145 (9)	0.139	0.166	0.149	-1.20

^aP = 0.05, interaction effect [breast cancer survivor (BCS) vs non-cancer control (CON)].

^bP < 0.05, effect of phase [post-training (POST) vs pre-training (PRE)].

A linear mixed model using a group (BCS vs CON) × training (PRE vs POST) approach was taken to assess training-induced changes. BCS: Breast cancer survivor; CON: Non-cancer control; PRE: Pre-training; POST: Post-training; CPET: Cardiopulmonary exercise test; VO₂peak: Peak oxygen uptake; G × P: Group × phase interaction; g: Hedge’s g effect sizes with thresholds set at 0.16 (small), 0.38 (medium), and 0.76 (large).

Exercise training

At POST, both groups increased CPET duration by 8.5% (BCS: +113 ± 135 seconds; CON: +30 ± 69 seconds; P = 0.006; Table 2). VO₂peak and peak HR did not change; however, a greater increase in peak wattage (+11 Watts, 95%CI: 9-13 Watts, g = 0.34, P = 0.05) was reported for BCS compared to CON. Overall, the absolute wattage performed during the acute exercise trial (visit 3) increased by 9% at POST (BCS: +9 ± 4 W; CON: +5 ± 6 W; P < 0.001).

Acute and training-induced responses in epinephrine

To determine if the exercise-induced adrenergic response was altered between BCS and CON (Aim 1), the group x time interaction was examined prior to training (PRE). No group × time interaction was reported (P = 0.972). There was, however, a time effect, with a 4.75-fold increase in EPI at 0 hour (+87.4 pg∙mL^-1, 95%CI: 44.4-130.4 pg∙mL^-1, P < 0.001, Figure 1A) which returned to resting levels at 1 hour with no group effect. Next, we explored the training-induced (group × phase) EPI response between BCS and CON at POST (Aim 2). The exercise-induced EPI response was separated into two phases: Initiation (0 hour-BASE) and recovery (1 hour-0 hour). No group × phase interactions were observed for EPI initiation (P = 0.642) or recovery (P = 0.976). There was a training effect, where at POST, a 68% lower EPI initiation [-37 pg∙mL^-1, 95%CI: (-73.4 to -0.5 pg∙mL^-1), P = 0.047; Figure 1B] and 72% higher EPI recovery [+39 pg∙mL^-1, 95%CI: (0.6-77.3 pg∙mL^-1), P = 0.047; Figure 1B] was observed compared to PRE. To further explore this POST EPI response, Total EPI AUC was examined, with no group × phase interaction (P = 0.614) identified. However, support for the initiation and recovery effects previously described are suggested by POST EPI AUC observations [-1660 pg∙minute∙mL^-1, 95%CI: (-3385-64.3 pg∙minute∙mL^-1), P = 0.058, g = 0.51; Figure 1C] which demonstrated a moderate effect size despite failing to reach statistical significance.

Figure 1 The acute exercise-induced response of epinephrine before and after completing a 16-week community-based exercise training program. A: Before training (PRE) responses were assessed using a group (BCS vs CON) × time [baseline (BASE), 0 hour, 1 hour] linear mixed model approach; B: After training (POST) assessments were assessed using a group (BCS vs CON) × training (PRE vs POST) linear mixed model approach; C: Confirmation of training-induced epinephrine (EPI) responses were pursued using an unplanned post hoc area under the curve analysis of EPI at PRE and POST. Data are mean (SD) with post-hoc comparisons performed using Tukey’s HSD test. ^aP < 0.001 effect of time (0 hour vs BASE); ^bP < 0.05 effect of phase (POST vs PRE). BASE: Baseline; 0H: Immediately post-exercise; 1H: 1-hour post-exercise; BCS: Breast cancer survivor; CON: Non-cancer control.

Acute and training-induced responses in norepinephrine

At PRE, there was no group × time interaction (Aim 1) for NE (P = 0.220; Figure 2A). Similar to EPI, there was a main effect of time, with a 4.85-fold increase in NE at 0 hour [+1295 pg∙mL^-1, 95%CI: (947-1643 pg∙mL^-1), P < 0.001] that returned to resting levels by 1 hour. However, a 35% lower NE response was observed in BCS at 0 hour [-575.9 pg∙mL^-1, 95%CI: (-1157 to 5.3 pg∙mL^-1),P = 0.350, g = 0.72] that had a borderline large effect size but was not statistically significant. Albeit nonsignificant, a large effect size was reported for the group × phase interaction at POST [-544.9 pg∙mL^-1, 95%CI: (-1214 to 124 pg∙mL^-1), P = 0.115, g = 0.92], suggesting the response divergence observed may be clinically meaningful if appropriately powered (Figure 2B). From PRE to POST, a 28% increase in NE initiation (0 hour-BASE) was observed for BCS [+341 pg∙mL^-1, 95%CI: (-163 to 845 pg∙mL^-1), P = 0.175, g = 0.36] compared to a 13% reduction for CON [-204 pg∙mL^-1, 95%CI: (-673 to 265 pg∙mL^-1), P = 0.377, g = 0.23]. Alternatively, no group × phase interaction was observed for NE recovery (P = 0.194). Further post hoc assessment via Total NE AUC analysis resulted in a non-significant group × phase interaction (P = 0.356; Figure 2C), which failed to confirm the previously described observations for NE initiation.

Figure 2 The acute exercise-induced response of norepinephrine before and after completing a 16-week community-based exercise training program. A: Before training (PRE) responses were assessed using a group (BCS vs CON) × time (BASE, 0 hour, 1 hour) linear mixed model approach; B: After training (POST) assessments were assessed using a group (BCS vs CON) × training (PRE vs POST) linear mixed model approach; C: Confirmation of training-induced norepinephrine (NE) responses were pursued using an unplanned post hoc area under the curve analysis of NE at PRE and POST. Data are mean (SD) with post-hoc comparisons performed using Tukey’s HSD test. ^aP < 0.001 effect of time (0 hour vs BASE). BASE: Baseline; 0H: Immediately post-exercise; 1H: 1-hour post-exercise; BCS: Breast cancer survivor; CON: Non-cancer control.

Correlations between training-induced catecholaminergic and immune cell responses

To determine if an altered exercise-induced adrenergic response may explain the previously reported attenuation (Aim 1) and training-induced recovery (Aim 2) of immune cell mobilization, Pearson correlations were explored to assess the association between exercise-induced immune cell activity and catecholaminergic initiation (0 hour-BASE) and recovery (1 hour-0 hour). As illustrated in Table 4, exercise-induced EPI initiation only accounted for 11% of response variability in lymphocyte mobilization [r = 0.34, 95%CI: (0.07-0.56), P = 0.016], whereas more moderate correlations were observed for NE initiation between lymphocyte [r = 0.53, 95%CI: (0.29-0.70), P < 0.001] and neutrophil [r = 0.54, 95%CI: (0.31-0.72), P < 0.001] mobilization, potentially contributing to 28% and 30% response variance, respectively. Alternatively, weak [r = 0.38, 95%CI: (0.11-0.59), P = 0.007] and moderate [r = 0.47, 95%CI: (0.22-0.66), P < 0.001] correlations were reported between lymphocyte egress and the recovery of EPI and NE, respectively. All other outcomes were non-significant.

Table 4 Correlation matrix between exercise-induced responses in catecholamines and immune cells before and after training.

Catecholamines	Response phase	Immune cell type	r	r2	95%CI	P value
Epinephrine (pg∙mL-1)	Initiation	Lymphocyte mobilization	0.34	0.11	(0.07-0.56)	0.016a
	Recovery	Lymphocyte egress	0.38	0.14	(0.11-0.59)	0.007a
	Initiation	MAIT cell mobilization	0.10	0.01	(-0.20 to 0.38)	0.506
	Recovery	MAIT cell egress	0.13	0.02	(-0.18 to 0.41)	0.413
	Initiation	Neutrophil mobilization	0.11	0.01	(-0.18 to 0.38)	0.467
	Recovery	Neutrophil egress	0.16	0.03	(-0.13 to 0.43)	0.278
Norepinephrine (pg∙mL-1)	Initiation	Lymphocyte mobilization	0.53	0.28	(0.29-0.70)	0.001a
	Recovery	Lymphocyte egress	0.47	0.22	(0.22-0.66)	0.001a
	Initiation	MAIT cell mobilization	0.16	0.03	(-0.14 to 0.43)	0.300
	Recovery	MAIT cell egress	0.21	0.05	(-0.09 to 0.48)	0.166
	Initiation	Neutrophil mobilization	0.54	0.30	(0.31-0.72)	0.001a
	Recovery	Neutrophil egress	0.19	0.03	(-0.11 to 0.45)	0.219

^aP ≤ 0.05 significant correlation (r).

Pearson’s correlations were assessed to investigate the association between training-induced changes in catecholaminergic and immune cell responses post-exercise. r: Pearson’s r with thresholds set at weak (< 0.40), moderate (0.40-0.69), and strong (≥ 0.70); r²: Coefficient of determination; MAIT: Mucosal-associated invariant T-cells.

DISCUSSION

Exercise training is beneficial for alleviating the adverse side effects of anti-cancer treatments. Our group has previously demonstrated the training-induced revival of immune cell activity in BCS[23,24], of which the mechanism underpinning this response is unknown. As such, we compared the acute exercise-induced EPI and NE responses of BCS and CON before and after a 16-week training intervention. Contrary to our hypotheses, our results demonstrated relatively similar exercise-induced catecholaminergic responses at PRE and POST between BCS and CON, without evidence of basal sympathoadrenal dysfunction observed in BCS. Moreover, no clear associations emerged between exercise-induced catecholaminergic responses and immune cell activity post-training. This suggests a mechanism alternative to catecholaminergic signaling may drive the training-induced immune cell recovery previously observed[23,24].

To contextualize these findings, the limitations and strengths are presented first. At approximately 70%, session adherence was lower than other supervised exercise interventions in BC patients (approximately 83%) and survivors (approximately 85%)[35], which may have left participants short of meeting physical activity guidelines (> 150 minute∙week^-1)[28]. Although no group differences in adherence were identified, CON demonstrated higher exercise compliance to prescribed exercises. This imbalance in exercise exposure may have reduced the between-group contrast in training effects and contributed to the absence of significant training-induced outcomes. Nonetheless, the maintenance of catecholaminergic signaling in trained BCS may suggest a positive effect. However, variability due to tumor status and treatment type[36], as well as the lower overall training dose, should be considered when interpreting these results. Due to concerns of participant burden and survey fatigue within the overall investigation (ClinicalTrials.gov ID: NCT03760536), BASE measurements of chronic stressor exposure were not collected. This hinders the ability to draw conclusions about the presentation of stress-induced adrenergic dysfunction in the BCS cohort[2-4], although the lack of BASE differences at PRE suggests this likely was not an issue. Finally, it is important to note that the original power calculations for this study were based on detecting changes in aerobic fitness (VO₂peak) as the primary outcome of interest[30]. As a result, the present secondary analysis which examined catecholaminergic responses may be underpowered to detect significant differences, as evidenced by the several borderline significant group × training interactions observed with moderate-to-large effect sizes. Future studies with adequate statistical power are warranted to confirm these observations and better define their clinical relevance for exercise prescription in cancer survivors. However, our study’s strength lies in the novel approach employed for biomarker investigation. By challenging the catecholaminergic response with an acute exercise bout performed at PRE and POST, an appropriate stimulus was provided to initiate sympathoadrenal stimulation that cannot be observed at rest[37-39]. Additionally, the inclusion of cancer-free, age- and body composition-matched females strengthened our methodology by providing a comparable referent group reflective of a ‘normal’ physiological response while avoiding ethical concerns that could arise with care randomization for BCS.

In the current study, naïve exercise-induced catecholaminergic responses did not differ between BCS and CON, with both groups exhibiting similar increases in EPI and NE immediately post-exercise before returning to BASE within one hour. Similar acute exercise-induced increases in catecholaminergic levels have been previously reported in BC patients[17] and survivors[36]; however, those studies lacked non-cancer female comparisons. In contrast, studies which included control groups reported blunted EPI responses in BCS[20] and PCa patients[21]. Although unexpected[27], the attenuated EPI responses observed in prior BCS studies may be attributed to differences in exercise intensities, as higher intensities generally elicit a more robust sympathoadrenal response[40-42]. For instance, Evans and colleagues[20] used a similar intermittent cycling protocol performed at approximately 60% VO₂peak (approximately 72% HRmax), which was a lower intensity compared to the approximately 60% PPO (approximately 85% HRmax) used within the current study. Notably, the intensity employed by Evans et al[20] falls below the ≥ 60% VO₂max (approximately 75% HRmax) intensity threshold previously shown to reliably produce a strong sympathoadrenal response[27,31], further demonstrating the potential role of exercise intensity in the differential catecholaminergic responses observed between BCS cohorts. Additionally, Evans et al[20] reported a 1.8-fold higher basal EPI measurement in BCS compared to the control group, which may have contributed to the attenuated exercise-induced response observed and could reflect elevated BASE stress levels, as previously reported in BCS[2,3]. Conversely, the current study found no group differences in basal EPI or NE, potentially suggesting the BCS cohort did not exhibit elevated levels of treatment-related[13,14] or stressor-induced[2-4] adrenergic dysfunction, as previously described. Discrepancies between EPI responses observed here and those reported in PCa survivors remain unclear, especially given the methodological similarities and comparable NE responses at exercise cessation[21]. Albeit speculative, potential modulators such as treatment history[12], sex[27], and cancer type[43] may influence catecholaminergic activity and warrant further investigation. Nevertheless, the absence of group differences in naïve catecholaminergic responses suggests a limited role for these hormones in explaining the previously observed attenuated exercise-induced immune mobilization in BCS[23,24].

At POST, evidence of a training effect on exercise-induced catecholaminergic responses was limited. Although not significant, training-induced reductions in EPI (-1660 pg∙minute∙mL^-1, P = 0.058) demonstrated a moderate effect size (g = 0.51), suggesting the possibility of training-induced autonomic adaptations not clearly captured due to limitations in sample size or variability in individual responses[27,44]. A similar nonsignificant interaction effect was observed for NE initiation (-544.9 pg∙mL^-1, P = 0.115), accompanied by a large effect size (g = 0.92), with BCS exhibiting a 28% increase and CON a 13% decrease in post-exercise NE from PRE to POST. While these divergent trends in NE may reflect emerging group-specific adaptations in training-induced autonomic efficiency, interpretations should remain limited given the lack of statistical significance and supporting evidence from AUC analyses. Although prior studies have reported attenuations in catecholaminergic responses consistent with adaptations such as enhanced clearance or reduced secretion rates[27], our findings remain inconclusive. Similarly, evidence remains limited in BCS, and to our knowledge, no studies have concurrently integrated initiation and recovery stimuli within the training response. In parallel analyses of immune activity, moderate but consistent correlations emerged between catecholaminergic and lymphocyte activity, yet no associations were reported for MAIT cells despite their lymphocyte classification and AR expression[45]. A moderate correlation was also observed between NE initiation and neutrophil mobilization, suggesting support for reports of NE-enhanced neutrophil phagocytosis post-exercise[46,47] and our group’s previous findings in trained BCS[23]. Collectively, these inconsistencies may be influenced by factors such as: Body composition[48], sex differences[44], synthetic origin of catecholamine release[49], and methodological limitations[27], underscoring the need for further investigation. Ultimately, the presence of training-induced catecholaminergic adaptations and their role in immune cell activation remains unclear, suggesting a potential disconnect within the exercise-induced neuroimmune connection of trained BCS.

Regardless, the current findings hold several important implications for future research and practical applications of exercise interventions within the management of BCS care. Most notably, the proposed maintenance of BCS catecholaminergic activity suggests that exercise training may contribute to the preservation or restoration of autonomic function, which is often compromised by anti-cancer treatments[13,14]. Although our results do not further clarify the catecholaminergic role in exercise-induced immune cell modulation, observations of training-related catecholaminergic maintenance - paired with functional performance improvements (e.g., CPET peak power output, exercise duration) - supports the efficacy of community-based exercise programs in this population. These interpretations are strengthened by intentional methodological decisions, particularly the inclusion of: (1) A referent control group; and (2) An appropriate exercise-based stimulus to assess training-induced stress responsiveness. The inclusion of CON provided essential context as a physiological benchmark, enabling more accurate evaluations of whether BCS responses reflect typical or atypical adaptations. Studies which exclude such comparators risk misinterpreting outcomes that may appear normal in isolation but deviate from expected exercise response patterns. Furthermore, the implementation of exercise challenges both before and after training offered a novel strategy to evaluate biological indices of training-induced stress, which have typically been assessed while at rest. Future investigations with mechanistic aims should consider adopting this approach to support physiologically valid interpretations. Lastly, the current findings emphasize the value of training compliance, as consistent adherence to the prescribed training regimen, even with lower session compliance, appeared sufficient to sustain positive adaptations in previously untrained BCS. However, the observed between-group differences in compliance underscore the need for future research to optimize intervention delivery and incorporate support mechanisms that enhance both exercise compliance and physiological responsiveness. Accordingly, clinical programs should prioritize strategies which foster participant engagement, motivation, and adaptive capacity, rather than focusing solely on adherence metrics.

CONCLUSION

In conclusion, the catecholaminergic role within the cancer care continuum remains a poorly understood complexity[8]. While chronically elevated catecholamine levels have been linked to adverse BC outcomes[9,10], acute exercise-induced catecholaminergic activity may confer protective effects through immune cell modulation[15-17]. Surprisingly, the present findings did not reveal catecholaminergic differences between BCS and CON, either before or after the training intervention. These results contrast with prior reports in cancer survivors[20,21] and suggest that alternative mechanisms may underlie the exercise-induced improvements in immune cell activity previously reported in trained BCS[23,24]. To further advance this area of work, future studies should incorporate acute physiological stressors surrounding training interventions to examine the capacity of systems to respond to stimuli. Identifying individuals with diminished physiological reserves, potentially placing them at a greater vulnerability for sympathoadrenal burnout, may help inform more effective exercise prescription in cancer survivors. Despite the absence of definitive evidence, the maintenance of sympathoadrenal activity underscores the safety and therapeutic viability of a community-based, moderate-intensity training program in BCS. Where feasible, future exercise interventions might explore the practicality of implementing individualized, evidence-based training, such as progressive intensity adjustments or interval-based approaches, as a means to optimize neuroendocrine adaptation and recovery[50,51]. These findings highlight the need for integrative research to clarify how cancer survivors differ in exercise-induced neuroendocrine and immune function, which can then guide the development of more targeted exercise prescriptions to mitigate cancer-related side effects.

ACKNOWLEDGEMENTS

The authors would like to thank the participants for their efforts during testing and training, as well as the efforts of Michael Bass, William Evans, Kaileigh Moertl, Cameron Stopforth, and Dean Amatuli for their assistance with data collection and administering the exercise intervention.